CN110264472B - SD-OCT-based fruit hypodermal cell nondestructive imaging method - Google Patents

SD-OCT-based fruit hypodermal cell nondestructive imaging method Download PDF

Info

Publication number
CN110264472B
CN110264472B CN201910458419.4A CN201910458419A CN110264472B CN 110264472 B CN110264472 B CN 110264472B CN 201910458419 A CN201910458419 A CN 201910458419A CN 110264472 B CN110264472 B CN 110264472B
Authority
CN
China
Prior art keywords
image
cell
fruit
diameter
oct
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201910458419.4A
Other languages
Chinese (zh)
Other versions
CN110264472A (en
Inventor
周扬
汪凤林
吴迪
周武杰
叶绿
石龙杰
岑岗
施秧
刘铁兵
陈正伟
陈才
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang Lover Health Science and Technology Development Co Ltd
Original Assignee
Zhejiang Lover Health Science and Technology Development Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhejiang Lover Health Science and Technology Development Co Ltd filed Critical Zhejiang Lover Health Science and Technology Development Co Ltd
Priority to CN201910458419.4A priority Critical patent/CN110264472B/en
Publication of CN110264472A publication Critical patent/CN110264472A/en
Application granted granted Critical
Publication of CN110264472B publication Critical patent/CN110264472B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/155—Segmentation; Edge detection involving morphological operators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
    • G06V20/693—Acquisition

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Multimedia (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)

Abstract

The invention discloses a fruit hypodermal cell nondestructive imaging method based on SD-OCT. The SD-OCT instrument is used for calibration, a plurality of continuous sectional images of the fruit subcutaneous cells are collected in a volume scanning mode, a three-dimensional data set is constructed, and each sectional image is compressed; smoothing the image and then combining gray threshold with binarization processing; dividing the region by using a watershed algorithm, and classifying the connected voxel points into a primary cell region; performing morphological closing operation on the preliminary cell area; and (3) calculating the morphological parameters of the three-dimensional space of each preliminary cell area, and determining cell screening conditions through microscopic images for screening. The invention realizes fruit near-subcutaneous cell imaging and three-dimensional morphological parameter statistics, has universality for the near-subcutaneous cell imaging and the three-dimensional morphological parameter statistics of different fruits, interactively selects a cell area, improves the accuracy of the parameters and lays a technical foundation for the fruit near-subcutaneous cell morphological analysis.

Description

SD-OCT-based fruit hypodermal cell nondestructive imaging method
Technical Field
The invention relates to a nondestructive imaging method for fruit hypodermal cells, in particular to a nondestructive imaging method for three-dimensional forms of fruit hypodermal cells based on an SD-OCT (secure digital-optical coherence tomography) technology.
Background
At present, the cell morphology observation mainly adopts a method of obtaining the cell morphology through biological section microscopic photographing and image processing, and also adopts a method of obtaining the cell morphology through a cell electronic imaging system based on a computer and corresponding software. Also, a method of preparing a cell suspension and measuring the morphology of cells in a cell counter. The first two methods include complex processes of fixing and staining fruit slices, and the use of reagents is highly required. Meanwhile, the original survival state of the cells is destroyed by the methods, and the original state of the cells is changed. The SD-OCT is adopted to scan the area to be detected of the fruit, so that nondestructive imaging is performed on the fruit hypodermal cells, and the method has the remarkable advantages of convenience, no damage, high efficiency and the like. On the premise of nondestructive detection, the three-dimensional morphological parameters of the cells can be accurately counted and analyzed, and the method has great significance in the research direction of cell morphology. When the fruit is damaged early, the cell morphology parameters of the two-dimensional RGB image obtained by photomicrography can not be used as the most appropriate parameters to reflect the change of the cells, so that the fruit hypodermal cell nondestructive imaging method based on the SD-OCT technology has a wide application prospect in the aspects of nondestructive detection of three-dimensional morphology of fruit hypodermal cells and the like.
Disclosure of Invention
In order to embody cells in three dimensions and carry out better imaging, the invention provides a fruit subcutaneous cell nondestructive imaging method based on SD-OCT.
The method comprises the steps of collecting multiple continuous section images of fruit hypodermal cells, filtering scattered noise in the images, carrying out image binarization, segmenting the binary images by a watershed algorithm to obtain a preliminary cell area, measuring morphological parameters of the preliminary cell area, reserving the preliminary cell area meeting cell size screening conditions, and finishing nondestructive imaging of the fruit hypodermal cells.
As shown in fig. 1, the specific technical solution adopted by the present invention is:
1) calibrating the SD-OCT instrument, and processing and setting theoretical resolution in each direction of the SD-OCT instrument in a volume scanning mode;
2) setting an SD-OCT instrument into a volume scanning mode, collecting a plurality of continuous section images of fruit subcutaneous cells, constructing and forming a three-dimensional data set, and compressing each section image according to the resolution in each direction; three-dimensional data are concentrated, and fruit images are processed and divided into three areas, namely a background area, a skin area and a pulp area;
3) smoothing the image, and reducing the scattered noise in the image;
4) converting the gray level image into a binary image by using the selected gray level threshold value and combining binarization processing;
in the implementation, the gray threshold is selected based on the comparison between the integrity of the cell region and the outer skin region being optimal.
5) Performing region segmentation on the binary image by using a watershed algorithm, classifying connected voxel points in the segmented image into a primary cell region, and concentrating a three-dimensional data to obtain a plurality of primary cell regions;
6) performing morphological closing operation on the preliminary cell region, filling holes inside the region, smoothing the boundary outside the region and eliminating small noise points;
7) calculating morphological parameters of a three-dimensional space of each preliminary cell Area, wherein the morphological parameters comprise barycenter coordinates in the Y direction, 3D surface Area3D, 3D Volume3D, equivalent sphere 3D diameter Eqdiameter3D, longest 3Dferet diameter Length3D and shortest 3Dferet diameter Width 3D; determining cell screening conditions through microscopic images, reserving a primary cell area which meets the screening conditions of cell size and has the gravity center in a cell layer, screening a final cell area in the primary cell area, and finishing the nondestructive imaging of the fruit hypodermal cells.
The step 1) is specifically as follows:
1.1) setting the SD-OCT instrument to be in a body scanning mode, then respectively shooting and collecting SD-OCT images of a glass plate, a glass slide and a glass ball according to the following modes, and further respectively setting X, Y, Z direction theoretical resolutions:
1.1.1) shooting and scanning a glass plate which is vertically arranged and has a known thickness along the horizontal direction in a volume scanning mode to obtain an SD-OCT image of the glass plate, wherein each image displays the left boundary and the right boundary of the glass plate, the number of voxel points between the two side boundaries (the left boundary and the right boundary) of the glass plate in the horizontal direction (X direction) is measured, and the X-direction theoretical resolution gamma of the SD-OCT is obtained by adopting the following formulaX：
Figure BDA0002077306490000021
1.1.2) shooting and scanning a glass slide which is vertically arranged and has a known thickness along the horizontal direction in a body scanning mode to obtain an SD-OCT image of the glass slide, wherein each image displays the upper boundary and the lower boundary of the glass slide, the number of voxel points between two side boundaries (the upper boundary and the lower boundary) in the vertical direction (Y direction) of the glass slide is measured, and the Y-direction theoretical resolution gamma of the SD-OCT is obtained by adopting the following formulaY：
Figure BDA0002077306490000022
Optical thickness-slide thickness x refractive index
1.1.3) shooting and scanning a glass ball which is embedded in epoxy resin glue and has a known diameter along the horizontal direction in a volume scanning mode to obtain an SD-OCT image of the glass ball, wherein each image displays the upper boundary and the lower boundary of a glass slide, the number of voxel points occupied by the top layer SD-OCT image and the bottom layer in the depth direction (Z direction, namely perpendicular to the X direction and the Y direction) of the glass ball is measured, and the Z-direction theoretical resolution gamma of the SD-OCT is obtained by adopting the following formulaZ：
Figure BDA0002077306490000031
The glass plate can be selected as a glass slide, the glass sheet can be selected as a cover glass, and the epoxy resin is glue and can be directly poured on the periphery of the glass ball.
Specifically, the measurement of the number of voxel points occupied by the top image and the bottom image of the glass ball (in the Z direction) can be scanned (the total scanning is 256 images, the glass ball is possibly scanned from the 15 th image as the top image, and the scanning of the glass ball is finished as the bottom image after the 219 th image, so that the section images of 15 to 219 are scanned to the glass ball, and the number of the voxel points occupied by the top image and the bottom image can be scanned to the number of 204 glass balls.)
The step 2) is specifically as follows:
2.1) using the SD-OCT instrument calibrated in the step 1) to set a volume scanning mode to scan a fruit region to be detected, and collecting a plurality of continuous section images of fruit hypodermal cells to form a three-dimensional image data set;
2.2) measuring the refractive index n of the fruit hypodermis by using a refractometer, and calculating the actual resolution of the instrument in the fruit hypodermis according to the following formula by combining the theoretical resolution of the SD-OCT instrument obtained in the step 1) in each direction:
γfruit (X direction)＝γX
Figure BDA0002077306490000032
γFruit (Z direction)＝γZ
Wherein, γFruit (X direction)Representing the actual resolution of the instrument in the X-direction, gamma, in the fruit tissueFruit (Y direction)Representing the actual resolution of the instrument in the Y direction in the fruit tissue, gammaFruit (Z direction)Representing the actual resolution of the instrument in the Z-direction in fruit tissue;
2.3) calculating the compression factor of the image by combining the following formula, so that the spatial distances represented by the voxels in the direction X, Y, Z are consistent, and the actual resolution gamma of the instrument in the fruit tissue is consistent:
Figure BDA0002077306490000033
Figure BDA0002077306490000034
γ＝γfruit (Z direction)
2.4) compressing each section image in the three-dimensional image data set by using a bilinear interpolation algorithm according to the compression multiple of the step 2.3), wherein the voxels in the X, Y direction of the compressed image are as the following formula:
the number of voxels in the X direction of the compressed image is X-direction compression multiple times of the number of voxels in the X direction of the section image
And (3) compressing the compressed image Y-direction voxel number which is the compression multiple of the Y direction and the original voxel number of the section image Y direction.
And 3) specifically, smoothing the compressed section image by using three-dimensional median filtering, wherein the connection types selected in the three-dimensional median filtering are a 6-neighborhood connection type, an 18-neighborhood connection type and a 26-neighborhood connection type. A 6 neighborhood connected type is defined as voxels with a common face considered connected, an 18 neighborhood connected type is defined as voxels with at least one common edge considered connected, and a 26 neighborhood connected type is defined as voxels with at least one common vertex considered connected.
The step 7) is specifically as follows:
7.1) searching edge voxels of a single preliminary cell Area, counting the number of the searched edge voxels, and obtaining the surface Area3D of the single preliminary cell Area according to the following formula:
area3D ═ the number of surface voxels × γ2
7.2) calculating the barycenter of each preliminary cell region in the Y-direction according to the following formula:
Figure BDA0002077306490000041
wherein, yiThe coordinate value of the position of the ith voxel in the Y direction is represented, i represents different voxel serial numbers in the preliminary cell area, and gamma represents the actual resolution of the instrument in fruit tissues;
7.3) calculating the sum of all voxel volumes of each preliminary cell area according to the following formula to obtain the Volume3D of each preliminary cell area;
volume3D ═ sum of voxel volumes × γ3
7.4) combining 7.3) Volume3D for each preliminary cell region, the equivalent spherical diameter EqDiameter3D of a single preliminary cell region is obtained according to the following formula:
Figure BDA0002077306490000042
7.5) calculating the diameter Length by utilizing the number of voxels of each preliminary cell region on one diameter, and searching the maximum value and the minimum value of all the diameter lengths of the barycenter of each preliminary cell region in the Y direction as the longest Feret diameter Length3D and the shortest Feret diameter Width3D respectively;
feret diameter-the number of voxels on diameter x γ
7.6) determining from the microscopic image the cell screening conditions of step 7) are:
7.6.1) will know the actual side length a1The square microscale is placed under a microscope for microscopic photographing to obtain a two-dimensional RGB image of the square microscale;
7.6.2) performing graying, binaryzation, hole filling and opening operation on the two-dimensional RGB image of the square microscale in sequence, calculating the number of voxel points occupied by the square as the area of the square, and squaring the area of the square to obtain the side length of the square, namely the number a of pixels occupied by the side length of the square2The resolution of the microscope is defined as γ according to the following formulaDisplay device：
Figure BDA0002077306490000051
Wherein, a2Representing the number of pixels occupied by the side length of a square, a1Representing the actual side length of a square microscale;
7.6.3) preparing a slice of the fruit hypodermal tissue, placing the slice under a microscope with the same state parameters as those in the step 7.6.1), and taking a photomicrograph to obtain a two-dimensional RGB image of the fruit hypodermal tissue;
7.6.4) graying a two-dimensional RGB image of a fruit hypodermis, segmenting the image into different regions by adopting a watershed algorithm, counting the number of pixel points in each region, calculating the region Area (Area2D) of each region and the average Area of all the regions, and filtering out a non-cell region in the image, wherein the non-cell region refers to a region with the region Area 1.5 times larger than the average Area of the region, so as to obtain a cell region in a microscopic image;
the number of pixels num1 of the equivalent circle 2D diameter (EqDiameter2D), the number of pixels num2 of the longest 2 dfferet diameter (Length2D), and the number of pixels num3 of the shortest 2 dfferet diameter (Width2D) of each cell Area in the microscopic image were statistically calculated, and the equivalent circle 2D diameter, the longest 2 dfferet diameter, and the shortest 2 dfferet diameter and Area2D of each cell were calculated:
area2D is the number of regional pixels multiplied by gammaDisplay device
EqDiameter2D＝num1×γDisplay device
Length2D＝num2×γDisplay device
Width2D＝num3×γDisplay device
7.7) respectively carrying out statistical sorting on the plurality of 2D cell parameters obtained in the step 7.6), finding out the maximum value of the equivalent circle diameter of the cell, the maximum value of the longest feret diameter and the minimum value of the shortest feret diameter, and then taking the values as references of cell size screening conditions to carry out screening according to the following cell size screening conditions: the cell sphere equivalent circle 3D diameter (EqDiameter3D) is less than the maximum value of the cell equivalent circle diameter (EqDiameter2D), the longest 3 dfferet diameter (Length3D) is less than the maximum value of the longest 2 dfferet diameter (Length2D), and the shortest feret diameter (Width3D) is greater than the minimum value of the shortest feret diameter (Width 2D);
and reserving the primary cell area which simultaneously meets the screening conditions of the cell size and has the gravity center in the pulp area, and screening the final cell area in the primary cell area to finish the nondestructive imaging of the fruit hypodermal cells.
The invention realizes fruit near subcutaneous cell imaging and three-dimensional morphological parameter statistics, has universality for the near subcutaneous cell imaging and the three-dimensional morphological parameter statistics of different fruits, interactively selects a cell area, improves the accuracy of the parameters and lays a technical foundation for the fruit subcutaneous cell morphological analysis.
The invention has the following beneficial effects:
the invention calibrates the SD-OCT instrument, calculates the three-dimensional morphological parameters of the subcutaneous cells of the fruit by three-dimensional image processing, and has the advantages of no damage, rapidness and low cost.
The invention adopts an interactive threshold parameter selection method, automatically cuts and counts the three-dimensional morphological parameters of the cells, thereby not only ensuring the integrity of the cells, but also ensuring the accuracy of the three-dimensional morphological parameters of the cells.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
FIG. 2 is a scaled two-dimensional sectional image of the SD-OCT acquisition. (a) Glass plate image, (b) slide image, (c) glass ball image.
Fig. 3 is a graph of the scaling effect. (a) A glass plate, (b) a glass slide, (c) a glass ball.
FIG. 4 is an image of cherry tomato near subcutaneous cells taken with SD-OCT. (a) Original image, (b) compressed image.
FIG. 5 is a two-dimensional effect graph of the median filtering step of the near-subcutaneous cell image of cherry tomato according to the present invention. (a) Before and after filtering.
FIG. 6 is a three-dimensional effect diagram of the median filtering step of the near-subcutaneous cell image of cherry tomato according to the present invention. (a) Before and after filtering.
FIG. 7 is a diagram showing the effect of binarization of near-subcutaneous cell images of cherry tomatoes according to the invention. (a) Two-dimensional effect map, and (b) three-dimensional effect map.
FIG. 8 is a diagram illustrating the effect of the watershed segmentation image of the near-subcutaneous cell image of cherry tomato according to the present invention. (a) Two-dimensional effect map, and (b) three-dimensional effect map.
FIG. 9 is a diagram of the effect of the preliminary cell region of the watershed segmentation image of the near-subcutaneous cell image of cherry tomato according to the present invention. (a) Two-dimensional effect map, and (b) three-dimensional effect map.
FIG. 10 is a graph showing the effect of the cherry tomato near subcutaneous cell image closure operation of the present invention. (a) Two-dimensional effect map, and (b) three-dimensional effect map.
FIG. 11 is a diagram showing the effect of the final cell region of the near-subcutaneous cell image of cherry tomato according to the present invention. (a) Two-dimensional effect maps, and (b) - (d) three-dimensional effect maps.
FIG. 12 is a two-dimensional RGB image processing effect diagram of the microscale of the present invention. (a) Two-dimensional RGB image effect picture of a microscale, (b) graying effect picture, (c) binaryzation effect picture, (d) filling hole effect picture, and (e) opening operation effect picture.
FIG. 13 is a graph showing the effect of two-dimensional RGB image processing on near-subcutaneous tissues of cherry tomatoes according to the invention. (a) The method comprises the following steps of (1) two-dimensional RGB image effect map of near subcutaneous tissue of cherry tomato, (b) graying effect map, (c) watershed effect map, and (d) retained cell area effect map.
Detailed Description
The present invention will be described in detail with reference to the drawings and examples, and it should be understood that the specific examples are illustrative only and are not intended to limit the present invention.
As shown in fig. 1, the embodiment of the present invention and the implementation process thereof are as follows:
1) calibrating the SD-OCT instrument to obtain the theoretical resolution in each direction when the SD-OCT instrument is set to be in a volume scanning mode
1.1) Using an OQ LabScope type SD-OCT imager manufactured by Thorlabs, set to a volume scanning mode in which 10 glass plates, which are vertically placed and have a thickness of about 2000um, are scanned in the horizontal direction, each image can display the left and right borders of the glass plate as a two-dimensional sectional image, as shown in FIG. 2 (a). Make itThe number of voxel points occupied by the glass plate (X direction, horizontal direction), i.e., between the left and right boundaries, was measured using the probe in avizo, as shown in FIG. 3(a), and the X-direction resolution theory γ of SD-OCT was obtained using the following formulaXCalculating the average value of 10 groups of X-direction theoretical resolution as the final X-direction theoretical resolution gammaXThe results are shown in Table 1.
Figure BDA0002077306490000071
Table 110 sets of average values of theoretical resolution in the X direction
Figure BDA0002077306490000072
1.2) using an OQ LabScope type SD-OCT imager manufactured by Thorlabs, set to a volume scanning mode in which 10 SD-OCT image samples horizontally placed and having a thickness of about 1000um slide glass are scanned in the horizontal direction, each image can show the upper and lower boundaries of the slide glass, a two-dimensional sectional image, as shown in (b) of FIG. 2. The number of voxel points occupied between the upper and lower boundaries of the slide glass (Y direction, vertical direction) was measured using a probe in avizo, as shown in FIG. 3(b), and the refractive index of the slide glass was measured with a gemstone refractometer to be 1.52, and the theoretical resolution γ in the Y direction of SD-OCT was obtained using the following formulaYAnd calculating the average value of 10 groups of Y-direction theoretical resolutions as the final Y-direction theoretical resolution gammaYThe results are shown in Table 2.
Optical thickness-physical thickness x refractive index
Figure BDA0002077306490000081
TABLE 2.10 average values of the theoretical resolution in the Y-direction
Figure BDA0002077306490000082
1.3) set up to a volume scanning mode using OQ LabScope type SD-OCT imager manufactured by Thorlabs corporation, scan 10 glass spheres embedded in epoxy resin glue and having a diameter of about 2000um in the horizontal direction, two-dimensional sectional images, as shown in FIG. 2(c), measure the number of voxel points occupied by the glass spheres (Z direction), i.e., the initial image and the final image of the glass spheres by scanning using a probe in avizo, as shown in FIG. 3(c), and obtain the Z direction theoretical resolution γ of SD-OCT using the following formulaZCalculating the average value of 10 groups of Z-direction theoretical resolutions as the final Z-direction theoretical resolution gammaZThe results are shown in Table 3.
Figure BDA0002077306490000083
TABLE 3 average Z-direction resolution
Figure BDA0002077306490000084
2) The SD-OCT instrument is set to be in a volume scanning mode, a plurality of continuous section images of fruit subcutaneous cells are collected and constructed to form a three-dimensional data set, and each section image is compressed according to the resolution ratio of each direction.
2.1) scanning the region to be measured of the cherry tomatoes by using the SD-OCT instrument which is calibrated in the step 1) and is set to be in a volume scanning mode, acquiring 256 continuous section images of near-subcutaneous cells of the cherry tomatoes, wherein the number of voxel points in the X direction of each section image is 256, and as shown in fig. 4(a), the number of voxel points in the Y direction of each section image is 512, so as to form a three-dimensional image data set;
2.2) taking mixed juice from two ends of cherry tomatoes, filtering solid substances in the juice by using filter paper, and measuring the refractive index of the cherry tomatoes by using an Abbe refractometer to be 1.351. Combining the theoretical resolution of the SD-OCT instrument in each direction in the step 1), calculating the actual resolution of the instrument in the cherry tomato tissue and the actual resolution gamma of the instrument in the X direction in the cherry tomato tissue according to the following formulaCherry tomato (X direction)Is 9.276 um/voxel and is,actual resolution gamma of instrument in Y direction in cherry tomato tissueCherry tomato (Y direction)3.996 um/voxel, the actual resolution γ of the instrument in the Z direction in cherry tomato tissueCherry tomato (Z direction)9.269 um/voxel.
2.3) calculating the compression factor of the image by combining the following formula, so that the spatial distances represented by the voxels in the X, Y, Z direction are consistent, and compressing the Y direction of the sectional image by 1/1.93 times according to the X, Y, Z direction cherry tomato actual resolution result, as shown in fig. 4(b), so that the actual resolution gamma of the instrument in the cherry tomato tissue is 9.27 um/voxel.
2.4) compressing the image by using a bilinear interpolation algorithm according to the compression multiple of the step 2.3), wherein voxels in the X, Y direction of the compressed image are 256 voxels in the X direction of the original image, 512 voxels in the Y direction of the original image, 256 voxels in the X direction of the compressed image and 266 voxels in the Y direction of the compressed image.
3) Smoothing the image, reducing speckle noise in the image, a two-dimensional effect map, as shown in fig. 5(b), and a three-dimensional effect map, as shown in fig. 6 (b). And (4) carrying out two iterations on the compressed section image by using three-dimensional median filtering. And selecting a connection type, sorting the gray values of all voxels in the neighborhood from the minimum value to the maximum value, finding a median value in the sequence, and replacing all voxel values in the neighborhood with the median value.
4) The gray level threshold is set on the premise that the contrast between the integrity of the cell region and the outer skin region is optimal, and the gray level image is converted into a binary image by using the gray level threshold in combination with binarization, a two-dimensional effect map, as shown in fig. 7(a), and a three-dimensional effect map, as shown in fig. 7 (b).
By adopting an interactive threshold selection method, in the threshold adjustment process, when the contrast between the integrity of the cell region and the outer skin region is observed to be optimal, the selected range is 0-25, the voxel value with the gray threshold range of 0-25 is set as 1, and the voxel value outside the gray threshold range is set as 0, namely a binary image.
5) A distance transformation watershed algorithm is used for the binary image to complete region segmentation, and a two-dimensional effect map is obtained, as shown in fig. 8(a), and a three-dimensional effect map is obtained, as shown in fig. 8(b), connected voxel points in the segmented image are classified into a preliminary cell region, and a plurality of preliminary cell regions can be obtained in a three-dimensional data set, as shown in fig. 9(a), and a three-dimensional effect map is obtained, as shown in fig. 9 (b).
5.1) in the first step, a chamfering algorithm is adopted to complete a distance-to-distance image with distance change on a binary image, the image is scanned twice in the front and back direction through a chamfering template, the first scanning is from right to left, from top to bottom, and the scanning is from left to right and from bottom to top. Obtaining an initial seed point from the distance image, and combining redundant seed points to realize optimization by taking the spatial information of the seed points as a judgment criterion; and secondly, reconstructing a distance map according to the optimized seed points, and completing image segmentation by adopting a fast intrusion watershed algorithm to obtain a segmented image.
And 5.2) classifying the 26-field connected voxel points in the segmentation image into a primary cell area, wherein a plurality of primary cell areas can be obtained in a three-dimensional data set. The specific method comprises the following steps: in the image, the start position to the end position of a connected region are iteratively detected from top to bottom and from left to right, the connected voxel points are set to be the same value and are classified into a primary cell region, the initial cell region is set from a value of 1, and a plurality of primary cell regions are set to be different values.
6) Performing morphological closing operation on the preliminary cell region, filling holes inside the region, smoothing the boundary outside the region, eliminating small noise points, and obtaining a two-dimensional effect map, as shown in fig. 10(a), and a three-dimensional effect map, as shown in fig. 10 (b).
Defining the cube of 3 voxels by 3 voxels as an element of a structure A, moving the element A, coinciding the center of the element A with the voxel, setting the value of the voxel as the value of other voxels in the neighborhood of the element A, determining the value of each voxel through two iterations, filling holes inside each preliminary cell region in the three-dimensional data set in the process, smoothing the boundary outside the region, and eliminating small noise outside the region.
7) Calculating morphological parameters of the three-dimensional space of each preliminary cell Area, wherein the morphological parameters comprise barycenter coordinates in the Y direction, 3D surface Area3D, 3D Volume3D, equivalent sphere 3D diameter Eqdiameter, longest 3Dferet diameter Length3D and shortest 3Dferet diameter Width 3D; the screening conditions of the cell size are determined through microscopic images, the preliminary cell area meeting the screening conditions of the cell size and the preliminary cell area with the gravity center in the cell layer are reserved, the final cell area is screened in the preliminary cell area, and then the nondestructive imaging of the fruit hypodermal cells is completed. Two-dimensional effect maps, as shown in FIG. 11(a), and three-dimensional effect maps, as shown in FIGS. 11 (b-d).
7.1) searching edge voxels of a single preliminary cell Area, and counting the number of the searched edge voxels to obtain the surface Area3D of the single preliminary cell Area.
7.2) calculate the barycenter of each preliminary cell region in the Y-direction.
7.3) calculating the sum of all voxel volumes of each preliminary cell area to obtain the Volume3D of each preliminary cell area.
7.4) combining 7.3) Volume3D of each preliminary cell area, the equivalent spherical diameter EqDiameter of a single preliminary cell area is obtained according to the following formula.
7.5) calculating the diameter Length by utilizing the number of voxels of each preliminary cell area on one diameter, and searching the maximum value and the minimum value of all the diameter lengths of the barycenter Barycenter of each preliminary cell area in the Y direction as the longest Feret diameter Length3D and the shortest Feret diameter Width3D respectively.
7.6) screening conditions for determining cell size by photomicrography were the following steps:
7.6.1) placing a square microscale with side length a1 of 100um under microscope, adjusting coarse quasi-focal helix, fine quasi-focal helix and stage of microscope by using large aperture and objective lens with magnification of 10, taking the best standard of black-white contrast of square displayed in image acquisition window, and acquiring batch two-dimensional RGB images of microscale by multiple acquisition as shown in FIG. 12(a)
7.6.2) removing hue and saturation information of the image from the two-dimensional RGB image of the microscale while preserving brightness to convert the RGB image into a grayscale image, i.e., graying the two-dimensional RGB image of the microscale, as shown in FIG. 12 (b); using the maximum inter-class variance method to find out that the threshold of the gray image is 0.596, that is, the critical value of the pixel gray value is 0.596 × 255 equals to 152, so that the gray value of the pixel in the gray image with the gray value of the pixel greater than 152 is 255, and the gray value of the pixel in the gray image with the gray value less than or equal to 152 is 0, that is, the two-dimensional RGB image of the microscopic scale is binarized, as shown in fig. 12 (c); filling hole regions in the binary image, as shown in fig. 12 (d); the image after filling the hole takes an area of an approximate disc surrounded by 10 pixel points as a template, and morphological opening operation is carried out, namely image processing of a two-dimensional RGB image of a microscale is completed, and the image is shown in FIG. 12 (e);
for the image whose image processing has been completed, the number of pixels occupied by the square, that is, the area of the square is calculated, the area of the square is squared to obtain the side length of the square, that is, the number a2 of pixels occupied by the side length of the square, and the resolution γ of the microscope is calculated according to the following formulaDisplay deviceAs shown in Table 4, the resolution of the microscope was 0.85 um/pixel by calculating and counting the average of the two-dimensional RGB images on the batch microscope scale.
TABLE 4 resolution of the microscope
Figure BDA0002077306490000111
7.6.3) the area to be measured of the cherry tomatoes is perpendicular to the epidermal slices to obtain cherry tomato hypodermal tissue slices, the slices are sliced for multiple times, the thinnest slice is selected and placed in the center of a clean glass slide with a drop of clear water, the clean glass cover is clamped by tweezers, the glass cover is slowly put down from one side of the clear water to avoid generating bubbles, the slices of the cherry tomato hypodermal tissue are manufactured, the slices are placed under a microscope with the same parameters as those in the step 7.6.1), the images at the moment are photographed by taking the optimal contrast of cell dividing lines displayed in the collected image window as a criterion, and two-dimensional RGB images of the cherry tomato hypodermal tissue are obtained, and as shown in FIG. 13(a), a batch of two-dimensional RGB images of the cherry tomato hypodermal tissue are obtained through multiple collection.
7.6.4) removing hue and saturation information of the two-dimensional RGB image of near-subcutaneous tissue of cherry tomato while preserving brightness to convert the RGB image into gray image, i.e. graying the two-dimensional RGB image of near-subcutaneous tissue of cherry tomato, as shown in FIG. 13 (b); filtering the gray level image in the horizontal and vertical directions by using a sobel edge operator to obtain a gradient amplitude image, segmenting the image into different regions by adopting a watershed algorithm on the gradient amplitude image, counting the number of pixel points of each region as shown in fig. 13(c), calculating the Area of the region (Area2D) according to the following formula, filtering out acellular regions with the Area of the region being more than 1.5 times of the average Area in the image, counting the number of pixels num1 of an equivalent circle diameter (EqDiameter2D) of each cell as shown in fig. 13(d), counting the number of pixels num2 of a longest feret diameter (Length2D) and the number of pixels num3 of a shortest feret diameter (Width2D) of each cell, and calculating the equivalent circle diameter, the longest feret diameter and the shortest feret diameter of each cell according to the following formula. Two-dimensional RGB images of batch cherry tomato hypodermal tissues are obtained according to multiple collection, cell parameters of the batch are obtained through statistics, the maximum value of equivalent circle diameters, the maximum value of the longest feret diameter and the minimum value of the shortest feret diameter in cell parameter results are found through comparison, as shown in Table 5, the maximum value and the minimum value are used as references of cell size screening conditions, and the cell size screening conditions are as follows: the cell sphere equivalent circle diameter (Eqdiameter3D) is less than 307.7um, the longest feret diameter (Length3D) is less than 521.9um, and the shortest feret diameter (Width3D) is greater than 81.6 um.
The cell parameter results are given in table 5 below:
number of pixels num1 Maximum value of equivalent circle diameter (um)
362 307.7
Number of pixels num2 Longest feMaximum value of ret diameter (um)
614 521.9
Number of pixels num3 Minimum value of shortest feret diameter (um)
96 81.6
7.7) preliminary cell area that will satisfy the screening condition of cell size, cell layer is mainly in the position of 500um-900um in its center of gravity in the upper half area of the whole image, so the screening condition is: 500um < BarycenterY <900um & Eqdiameter3D <307.7um & Length3D <521.9um & Width3D >81.6um, and screening a final cell area in the primary cell area according to screening conditions, namely completing nondestructive imaging of near-subcutaneous cells of the cherry tomatoes.
In the embodiment of the present invention, it can be further understood by those skilled in the art that all or part of the steps in the method for implementing the above embodiment may be implemented by instructing the relevant hardware through a program, where the program may be stored in a computer-readable storage medium, where the storage medium includes a ROM/RAM, a magnetic disk, an optical disk, and the like.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.

Claims (4)

1. A fruit hypodermal cell nondestructive imaging method based on SD-OCT is characterized by comprising the following steps:
1) calibrating the SD-OCT instrument, and processing and setting theoretical resolution in each direction of the SD-OCT instrument in a volume scanning mode;
2) setting an SD-OCT instrument into a volume scanning mode, collecting a plurality of continuous section images of fruit subcutaneous cells, constructing and forming a three-dimensional data set, and compressing each section image according to the resolution in each direction; three-dimensional data are concentrated, and fruit images are processed and divided into three areas, namely a background area, a skin area and a pulp area;
3) smoothing the image, and reducing the scattered noise in the image;
4) converting the gray level image into a binary image by using a gray level threshold value in combination with binarization processing;
5) performing region segmentation on the binary image by using a watershed algorithm, classifying connected voxel points in the segmented image into a primary cell region, and concentrating a three-dimensional data to obtain a plurality of primary cell regions;
6) performing morphological closing operation on the preliminary cell area, filling holes inside the area and smoothing the external boundary of the area;
7) calculating morphological parameters of a three-dimensional space of each preliminary cell Area, wherein the morphological parameters comprise barycenter coordinates in the Y direction, 3D surface Area3D, 3D Volume3D, equivalent sphere 3D diameter Eqdiameter3D, longest 3Dferet diameter Length3D and shortest 3Dferet diameter Width 3D; determining cell screening conditions through microscopic images, reserving a primary cell area which meets the screening conditions of cell size and has the gravity center in a cell layer, and screening a final cell area in the primary cell area to finish the nondestructive imaging of the fruit hypodermal cells;
the step 1) is specifically as follows:
1.1) setting the SD-OCT instrument to be in a body scanning mode, then respectively shooting and collecting SD-OCT images of a glass plate, a glass slide and a glass ball according to the following modes, and further respectively setting X, Y, Z direction theoretical resolutions:
1.1.1) shooting and scanning a glass plate which is vertically arranged and has a known thickness along the horizontal direction in a body scanning mode to obtain an SD-OCT image of the glass plate, measuring the number of voxel points occupied between two side boundaries of the glass plate in the X direction, and obtaining the X direction theoretical resolution gamma of the SD-OCT by adopting the following formulaX：
Figure FDA0002902790280000011
1.1.2) shooting and scanning a glass slide which is vertically arranged and has a known thickness along the horizontal direction in a body scanning mode to obtain an SD-OCT image of the glass slide, measuring the number of voxel points occupied between two side boundaries of the glass slide in the Y direction, and obtaining the Y direction theoretical resolution gamma of the SD-OCT by adopting the following formulaY：
Figure FDA0002902790280000021
Optical thickness-slide thickness x refractive index
1.1.3) shooting and scanning a glass ball which is embedded in epoxy resin glue and has a known diameter along the horizontal direction under a volume scanning mode to obtain an SD-OCT image of the glass ball, measuring the top layer SD-OCT image and the number of voxels occupied by the bottom layer of the glass ball in the Z direction which is vertical to the X direction and the Y direction, and obtaining the Z direction theoretical resolution gamma of the SD-OCT by adopting the following formulaZ：
Figure FDA0002902790280000022
2. The SD-OCT based fruit hypodermal cell nondestructive imaging method of claim 1, wherein:
the step 2) is specifically as follows:
2.1) using the SD-OCT instrument calibrated in the step 1) to set a volume scanning mode to scan a fruit region to be detected, and collecting a plurality of continuous section images of fruit hypodermal cells to form a three-dimensional image data set;
2.2) measuring the refractive index n of the fruit hypodermis by using a refractometer, and calculating the actual resolution of the instrument in the fruit hypodermis according to the following formula by combining the theoretical resolution of the SD-OCT instrument obtained in the step 1) in each direction:
γfruit (X direction)＝γX
Figure FDA0002902790280000023
γFruit (Z direction)＝γZ
Wherein, γFruit (X direction)Representing the actual resolution of the instrument in the X-direction, gamma, in the fruit tissueFruit (Y direction)Representing the actual resolution of the instrument in the Y direction in the fruit tissue, gammaFruit (Z direction)Representing the actual resolution of the instrument in the Z-direction in fruit tissue; x-direction theoretical resolution gamma of SD-OCTXY-direction theoretical resolution γ of SD-OCTYZ-direction theoretical resolution γ of SD-OCTZ；
2.3) calculating the compression factor of the image by combining the following formula, so that the spatial distances represented by the voxels in the direction X, Y, Z are consistent, and the actual resolution gamma of the instrument in the fruit tissue is consistent:
Figure FDA0002902790280000024
Figure FDA0002902790280000031
γ＝γfruit (Z direction)
2.4) compressing each section image in the three-dimensional image data set by using a bilinear interpolation algorithm according to the compression multiple of the step 2.3), wherein the voxels in the X, Y direction of the compressed image are as the following formula:
the number of voxels in the X direction of the compressed image is X-direction compression multiple times of the number of voxels in the X direction of the section image
And (3) compressing the compressed image Y-direction voxel number which is the compression multiple of the Y direction and the original voxel number of the section image Y direction.
3. The SD-OCT based fruit subepithelial cell nondestructive imaging method of claim 2, wherein:
and 3) specifically, smoothing the compressed section image by using three-dimensional median filtering, wherein the connection types selected in the three-dimensional median filtering are a 6-neighborhood connection type, an 18-neighborhood connection type and a 26-neighborhood connection type.
4. The SD-OCT based fruit subepithelial cell nondestructive imaging method of claim 2, wherein: the step 7) is specifically as follows:
7.1) searching edge voxels of a single preliminary cell Area, counting the number of the searched edge voxels, and obtaining the surface Area3D of the single preliminary cell Area according to the following formula:
area3D ═ the number of surface voxels × γ2
7.2) calculating the barycenter of each preliminary cell region in the Y-direction according to the following formula:
Figure FDA0002902790280000032
wherein, yiThe coordinate value of the position of the ith voxel in the Y direction is represented, i represents different voxel serial numbers in the preliminary cell area, and gamma represents the actual resolution of the instrument in fruit tissues;
7.3) calculating the sum of all voxel volumes of each preliminary cell area according to the following formula to obtain the Volume3D of each preliminary cell area;
volume3D ═ sum of voxel volumes × γ3
7.4) combining 7.3) Volume3D for each preliminary cell region, the equivalent spherical diameter EqDiameter3D of a single preliminary cell region is obtained according to the following formula:
Figure FDA0002902790280000033
7.5) calculating the diameter Length by utilizing the number of voxels of each preliminary cell region on one diameter, and searching the maximum value and the minimum value of all the diameter lengths of the barycenter of each preliminary cell region in the Y direction as the longest Feret diameter Length3D and the shortest Feret diameter Width3D respectively;
feret diameter-the number of voxels on diameter x γ
7.6) determining from the microscopic image the cell screening conditions of step 7) are:
7.6.1) will know the actual side length a1The square microscale is placed under a microscope for microscopic photographing to obtain a two-dimensional RGB image of the square microscale;
7.6.2) performing graying, binaryzation, hole filling and opening operation on the two-dimensional RGB image of the square microscale in sequence, calculating the number of voxel points occupied by the square as the area of the square, and squaring the area of the square to obtain the side length of the square, wherein the resolution of the microscope is defined as gamma according to the following formulaDisplay device：
Figure FDA0002902790280000041
Wherein, a2Representing the number of pixels occupied by the side length of a square, a1Representing the actual side length of a square microscale;
7.6.3) preparing a slice of the fruit hypodermal tissue, placing the slice under a microscope with the same state parameters as those in the step 7.6.1), and taking a photomicrograph to obtain a two-dimensional RGB image of the fruit hypodermal tissue;
7.6.4) graying the two-dimensional RGB image of the fruit hypodermis, adopting a watershed algorithm to segment the image into different regions, counting the number of pixel points in each region, then calculating the region area of each region and the average area of all the regions, and filtering out the non-cell region in the image, wherein the non-cell region refers to a region with the region area 1.5 times larger than the average area of the region, so as to obtain the cell region in the microscopic image;
statistically calculating the number of pixels num1 of the equivalent circle 2D diameter, the number of pixels num2 of the longest 2 dfferet diameter and the number of pixels num3 of the shortest 2 dfferet diameter of each cell Area in the microscopic image, and further calculating the equivalent circle 2D diameter, the longest 2 dfferet diameter and the shortest 2 dfferet diameter and Area2D of each cell:
area2D is the number of regional pixels multiplied by gammaDisplay device
EqDiameter2D＝num1×γDisplay device
Length2D＝num2×γDisplay device
Width2D＝num3×γDisplay device
7.7) respectively carrying out statistical sorting on the plurality of 2D cell parameters obtained in the step 7.6), finding out the maximum value of the equivalent circle diameter of the cell, the maximum value of the longest feret diameter and the minimum value of the shortest feret diameter, and then taking the values as references of cell size screening conditions to carry out screening according to the following cell size screening conditions: the 3D diameter of the cell sphere equivalent circle is smaller than the maximum value of the cell equivalent circle diameter, the longest 3Dferet diameter is smaller than the maximum value of the longest 2Dferet diameter, and the shortest feret diameter is larger than the minimum value of the shortest feret diameter;
and reserving the primary cell area which simultaneously meets the screening conditions of the cell size and has the gravity center in the pulp area, and screening the final cell area in the primary cell area to finish the nondestructive imaging of the fruit hypodermal cells.
CN201910458419.4A 2019-05-29 2019-05-29 SD-OCT-based fruit hypodermal cell nondestructive imaging method Active CN110264472B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910458419.4A CN110264472B (en) 2019-05-29 2019-05-29 SD-OCT-based fruit hypodermal cell nondestructive imaging method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910458419.4A CN110264472B (en) 2019-05-29 2019-05-29 SD-OCT-based fruit hypodermal cell nondestructive imaging method

Publications (2)

Publication Number Publication Date
CN110264472A CN110264472A (en) 2019-09-20
CN110264472B true CN110264472B (en) 2021-05-04

Family

ID=67915866

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910458419.4A Active CN110264472B (en) 2019-05-29 2019-05-29 SD-OCT-based fruit hypodermal cell nondestructive imaging method

Country Status (1)

Country Link
CN (1) CN110264472B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111429440B (en) * 2020-03-31 2023-04-28 上海杏脉信息科技有限公司 Method, system, equipment, device and medium for detecting sufficiency of microscopic pathology image cells
CN111767809A (en) * 2020-06-18 2020-10-13 湖南理工学院 Intelligent cell identification method based on laser confocal microscopy
CN111932499B (en) * 2020-07-10 2023-12-22 深圳市瑞沃德生命科技有限公司 Cell diameter calculation method, device and system

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015017536A1 (en) * 2013-07-31 2015-02-05 The Board Of Trustees Of The Leland Stanford Junior University Method and system for evaluating progression of age-related macular degeneration
CN106558030A (en) * 2016-11-15 2017-04-05 苏州大学 The dividing method of three-dimensional big visual field frequency sweep optical coherence tomography median nexus film
CN106780347A (en) * 2017-02-09 2017-05-31 浙江科技学院 A kind of loquat early stage bruise discrimination method based on OCT image treatment
CN108257126A (en) * 2018-01-25 2018-07-06 苏州大学 The blood vessel detection and method for registering, equipment and application of three-dimensional retina OCT image

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106073788B (en) * 2016-07-19 2023-03-17 杭州捷诺飞生物科技有限公司 OCT-based in-situ three-dimensional printing skin repair equipment and implementation method thereof
CN109283202A (en) * 2017-07-21 2019-01-29 中国石油化工股份有限公司 It is a kind of for scanning the sample holder of subtle core

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015017536A1 (en) * 2013-07-31 2015-02-05 The Board Of Trustees Of The Leland Stanford Junior University Method and system for evaluating progression of age-related macular degeneration
CN106558030A (en) * 2016-11-15 2017-04-05 苏州大学 The dividing method of three-dimensional big visual field frequency sweep optical coherence tomography median nexus film
CN106780347A (en) * 2017-02-09 2017-05-31 浙江科技学院 A kind of loquat early stage bruise discrimination method based on OCT image treatment
CN108257126A (en) * 2018-01-25 2018-07-06 苏州大学 The blood vessel detection and method for registering, equipment and application of three-dimensional retina OCT image

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Deep Features and Data Reduction for Classification of SD-OCT Images: Application to Diabetic Macular Edema;Genevieve C. Y. Chan;《2018 International Conference on Intelligent and Advanced System (ICIAS)》;20181122;全文 *
谱域光学相干层析成像的深度拓展及其在眼调节研究中的应用;范姗慧;《中国博士学位论文全文数据库 医药卫生科技辑》;20170215;论文第23-25，76-77页 *

Also Published As

Publication number Publication date
CN110264472A (en) 2019-09-20

Similar Documents

Publication Publication Date Title
CN110264472B (en) SD-OCT-based fruit hypodermal cell nondestructive imaging method
CN109285222A (en) The building of organic shale high-resolution digital rock core and analysis method
KR100870412B1 (en) Ultrasound system for forming 3d fetus ultrasound image based on fetus surface image extracted by svm-based texture classification and method for the same
CN104751178B (en) Lung neoplasm detection means and method based on shape template matching combining classification device
CN107194872B (en) Remote sensed image super-resolution reconstruction method based on perception of content deep learning network
CN103907023B (en) Abnormal system and method in detection biological sample
CN109580630A (en) A kind of visible detection method of component of machine defect
CN108765438B (en) Liver boundary identification method and system
CN106023158B (en) The fresh water pipless pearl pearly layer defect identification method of SD-OCT images
US20060177125A1 (en) Computerized detection of breast cancer on digital tomosynthesis mammograms
CN102481117A (en) System And Method For Detecting Poor Quality In 3d Reconstructions
CN108186051A (en) A kind of image processing method and processing system of the automatic measurement fetus Double Tops electrical path length from ultrasonoscopy
CN110378875A (en) Internal lithangiuria ingredient discrimination method based on machine learning algorithm
CN116721391B (en) Method for detecting separation effect of raw oil based on computer vision
CN114565658B (en) CT technology-based pore size calculation method and device
CN108061697B (en) Method for calculating three-dimensional porosity of soil body
CN108378869A (en) A kind of image processing method and processing system of the automatic measurement fetal head girth degree from ultrasonoscopy
Maitra et al. Detection of abnormal masses using divide and conquer algorithmin digital mammogram
CN102369541B (en) Method for performing automatic classification of image information
CN108875741A (en) It is a kind of based on multiple dimensioned fuzzy acoustic picture texture characteristic extracting method
CN111524134B (en) Method and device for detecting regularity of honeycomb products on production line
CN105300302B (en) The measuring method of Brinell circular diameter
CN112991287A (en) Automatic indentation measurement method based on full convolution neural network
CN110348459A (en) Based on multiple dimensioned quick covering blanket method sonar image fractal characteristic extracting method
CN110490971B (en) Method for reconstructing cell dynamic characteristic three-dimensional image under biological microscope

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant